Metastatic medullary thyroid carcinoma: a new way forward

Medullary thyroid carcinoma (MTC) is a rare malignancy comprising 1–2% of all thyroid cancers in the United States. Approximately 20% of cases are familial, secondary to a germline RET mutation, while the remaining 80% are sporadic and also harbour a somatic RET mutation in more than half of all cases. Up to 15–20% of patients will present with distant metastatic disease, and retrospective series report a 10-year survival of 10–40% from time of first metastasis. Historically, systemic therapies for metastatic MTC have been limited, and cytotoxic chemotherapy has demonstrated poor objective response rates. However, in the last decade, targeted therapies, particularly multitargeted tyrosine kinase inhibitors (TKIs), have demonstrated prolonged progression-free survival in advanced and progressive MTC. Both cabozantinib and vandetanib have been approved as first-line treatment options in many countries; nevertheless, their use is limited by high toxicity rates and dose reductions are often necessary. New generation TKIs, such as selpercatinib or pralsetinib, that exhibit selective activity against RET, have recently been approved as a second-line treatment option, and they exhibit a more favourable side-effect profile. Peptide receptor radionuclide therapy or immune checkpoint inhibitors may also constitute potential therapeutic options in specific clinical settings. In this review, we aim to present all current therapeutic options available for patients with progressive MTC, as well as new or as yet experimental treatments.


Introduction
Medullary thyroid carcinoma (MTC) is a malignant neuroendocrine tumour originating from the parafollicular or C-cells of the thyroid, capable of secreting calcitonin and carcinoembryonic antigen (CEA) (Ceolin et al. 2019). It is a rare malignancy comprising 1-2% of all thyroid cancers in the United States (Tuttle et al. 2014, Lim et al. 2017. Approximately 20% of cases are familial, secondary to a germline rearranged during transfection (RET) mutation, while the remaining 80% are sporadic (Bai et al. 2020) and also harbour a somatic RET mutation in more than half of all cases, with RET M918T being the most frequent genetic alteration encountered in 30-50% of aggressive sporadic MTC (Dvorakova et al. 2008, Taccaliti et al. 2011.

Initial therapeutic approach to MTC
The standard treatment for MTC is total thyroidectomy and dissection of cervical lymph node compartments, as surgery is the only curative treatment (Wells et al. 2015). Patients with persistent or recurrent MTC localised to the neck are candidates for repeated neck explorations. However, in the presence of widespread regional or metastatic disease, extensive surgery is not associated with a higher cure rate or survival benefit and should be considered mainly for local symptom control (Wells et al. 2015, Randle et al. 2017. Post-operative biochemical remission of serum calcitonin is significantly correlated with an improved 5-year recurrence rate but not 5-year survival (Jung et al. 2016). The 5-year and 10-year survival rates are 25 and 8%, respectively, when the doubling time of calcitonin is less than 6 months, and 92 and 37%, respectively, and when the doubling time ranges from 6 months to 2 years (Wells et al. 2015). Currently, a short calcitonin and CEA doubling time (<6 months) are considered the best available indicators to assess tumour behaviour, MTC recurrence and cancer mortality (Barbet et al. 2005, Laure Giraudet et al. 2008. Post-operative levels of serum calcitonin > 150 ng/L generally indicate distant metastases (Laure Giraudet et al. 2008).
Neck and/or chest computerised tomography (CT), hepatic magnetic resonance imaging (MRI) and bone scintigraphy are the main imaging modalities utilised for MTC staging and follow-up (Maia et al. 2014, Wells et al. 2015. Although several radionuclide imaging modalities are available, positron emission tomography (PET) CT using 18 F-fluorodeoxyglucose ( 18 F-FDG), 18 F-DOPA and 68 Ga-DOTA-somatostatin analogues ( 68 Ga-SSA) offer higher sensitivity compared with conventional imaging (CT/MRI). Of interest, a study evaluating the performance of 68 Ga-1, 4,7,1 0-tet raaza cyclo dodec ane-1 ,4,7, 10-te traac etic acid-octreotate (DOTATATE)-PET/CT in detecting MTC lesions suggested that it is highly sensitive in identifying bone metastases and could be a substitute for bone scans, and importantly, as a theranostic, it indicates a potential therapeutic tool in the presence of considerable uptake (Castroneves et al. 2018, Hayes et al. 2021.

Metastatic disease
Up to 15-20% of patients will present with distant metastatic disease at diagnosis, and retrospective series report a 10-year survival of 10-40% from time of first metastasis (Roman et al. 2006, Chiacchiarini et al. 2021).
Cytotoxic chemotherapy, as well as local treatments such as external beam radiotherapy, has only a limited and short-term benefit (Wells et al. 2015). Currently available systemic treatments include the use of multityrosine kinase inhibitors (TKIs), which partially inhibit multiple kinases including RET, and often impair multiple signalling pathways (Filetti et al. 2019). TKIs have shown prolonged progression-free survival in patients with tumour progression, although at a cost of considerable toxicity (Elisei et al. 2013). Although the Food and Drug Administration (FDA) and European Medicines Agency (EMA) have approved several TKIs for treating advanced MTC, novel therapeutic modalities are still required to treat patients who fail to respond to TKIs or experience unacceptable toxicity. New-generation targeted therapies such as immunotherapy and peptide receptor radionuclide therapy (PRRT) may be promising but require prospective randomised trials.
In this review, we aim to present all current therapeutic options available for patients with progressive MTC, as well as the new or as yet experimental treatments based on the individualised molecular profile of each patient.

Therapeutic strategies for metastatic MTC
When evaluating a patient with advanced MTC, various factors need to be taken into account. Is the patient symptomatic or asymptomatic? Is the locoregional disease controlled? Is there any calcitonin-related secretory syndrome or ectopic adrenocorticotropic hormone (ACTH) syndrome? Are there lesions that require intervention due to an imminent risk or associated symptoms? What is the rate of progression (calcitonin doubling-time and imaging findings based on Response Evaluation Criteria in Solid Tumors (RECIST) criteria)?

Tyrosine kinase inhibitors
Advances in understanding the molecular mechanisms and intracellular signalling pathways involved in MTC pathogenesis have allowed for the development of targeted therapies, offering new perspectives on effective therapies for advanced MTC. Generally, these therapies are based on the blockade of the MAPK and RAS pathways, which promote MTC development. Many compounds have been developed to inhibit the activity of these pathways (Fig. 1). oncogenic pathways. Several multitargeted TKIs have been tested in MTC, including motesanib (Schlumberger et al. 2009), sorafenib (de Castroneves et al. 2016, sunitinib (Carr et al. 2010), axitinib (Cohen et al. 2008), imatinib (de Groot et al. 2007, pazopanib (Bible et al. 2014), anlotinib (Sun et al. 2018), lenvatinib (Haugen et al. 2016), vandetanib (Wells et al. 2012) and cabozantinib (Elisei et al. 2013) (Table 1). Of these various TKIs, only vandetanib and cabozantinib are currently FDA and EMA approved for treating unresectable and progressive advanced MTC (Wells et al. 2015. Because patients with progressive MTC may follow a relatively indolent disease course and can experience prolonged survival even without therapy, prospective trials should preferably mandate RECIST progression on trial enrolment and use progression-free survival (PFS) along with overall survival (OS) to evaluate drug efficacy (Hadoux & Schlumberger 2017).
The first approved TKI, vandetanib, selectively targets RET, vascular epithelial growth factor receptor (VEGFR) and EGF receptors (EGFR) (Wedge et al. 2002, Wells et al. 2012. The efficacy of vandetanib was evaluated in 331 individuals with progressive MTC (123 had tumour progression, either locoregional or distant metastases) randomised 2:1 to receive vandetanib (300 mg/day) or placebo (Zactima Efficacy in Thyroid Cancer Assessment (ZETA) study) (Wells et al. 2012). Inclusion criteria included measurable, unresectable locally advanced or metastatic, hereditary or sporadic MTC. The results showed a statistically significant increase in the PFS in the vandetanib-treated group compared to placebo (30.5 vs 19.3 months, P < 0.001), as well as a higher objective response rate. Vandetanib has also been successfully used to treat children with MEN2B (Fox et al. 2013).
The second approved TKI, cabozantinib (140 mg/day), is a c-MET, VEGFR2 and RET inhibitor. A randomised (2:1) study of 330 individuals demonstrated a significant increase in PFS in the cabozantinib-treated group compared to placebo (11.2 vs 4.0 months, P < 0.001) in the advanced MTC (EXAM) study (Elisei et al. 2013). In this trial, all patients had documented radiological disease progression according to RECIST (40% had received previous anticancer therapies, including other TKIs). The effect of vandetanib or cabozantinib on the overall survival of MTC patients remains unknown, but interim analyses have not revealed any difference between the experimental and placebo arms (Wells et al. 2012, Elisei et al. 2013). This could potentially be due to the cross-over design in the ZETA study, but patients receiving placebo were not allowed to cross-over in the EXAM study. In vivo data have shown Endocrine-Related Cancer  Endocrine-Related Cancer that both drugs inhibit angiogenesis in a dose-dependent manner, although cabozantinib showed more potent antiangiogenic activity than vandetanib (Carra et al. 2021). In a recent retrospective study including MTC patients (n = 48) with locally advanced disease and/or evidence of distant metastases treated with vandetanib and/or cabozantinib (as first-line and later-line treatment), the median PFS for first-line vandetanib-and cabozantinib-treated patients was 12 and 9 months, respectively. The median total OS (for the total duration of treatment, first-line and laterline settings) was 54 and 24 months, respectively (Koehler et al. 2021); however, these results are difficult to interpret given that significantly more patients received vandetanib in the first-line setting (85%) compared with cabozantinib (15%), and this likely accounts for the poorer prognosis of cabozantinib-treated patients. In vandetanib-treated patients, the PFS and OS were significantly longer in patients aged ≤60 years at TKI initiation and in patients with ≥5 treatment-related adverse events. In addition, the PFS was prolonged in those without bone metastases (Koehler et al. 2021). The genetic signature is crucial for predicting responsiveness to therapy in progressive MTC. Phase 3 trials have helped identify predictors of response to both vandetanib and cabozantinib. The subgroup of patients in the ZETA study with the RET M918T mutation showed a higher response rate to vandetanib than RET M918-negative patients (54.5% vs 32%), but these data were based on a small sample size (Wells et al. 2012). In a larger phase 3 trial, cabozantinib appeared to prolong PFS vs placebo in the RET mutation-positive subgroup of patients with radiologically documented progressive MTC (HR: 0.23; 95% CI: 0.14-0.38; P < 0.0001), in the RET mutation-unknown subgroup (HR: 0.30; 95% CI: 0.16-0.57; P = 0.0001) and in the RAS mutation-positive subgroup (HR: 0.15; 95% CI: 0.02-1.10; P = 0.0317). The RET M918T subgroup achieved the greatest observed PFS from cabozantinib compared to placebo (HR: 0.15; 95% CI: 0.08-0.28; P < 0.0001). Furthermore, the PFS following cabozantinib was prolonged in patients with RET M918T (61 vs 17 weeks) or RAS mutations compared to patients with other mutations . A microRNA expression analysis in MTC has shown increased expression of miR-375 and decreased protein levels of the miR-375 target SEC23A in MTC, both of which conferred increased sensitivity to vandetanib. In vitro studies have shown that RET codon 804 and 806 mutations confer resistance to vandetanib therapy (Carlomagno et al. 2009).
Common adverse effects of TKIs include hypertension, palmar-plantar erythrodysaesthesia, proteinuria and QTc interval prolongation, with the most prominent being fatigue, lethargy, anorexia and gastrointestinal disturbances; however, all TKIs have a unique adverse event profile (Table 1). Approximately 35% of the patients treated with vandetanib had a QTc prolongation of more than 60 ms, which is considered a grade 4 adverse event requiring dosage reduction, but this was not encountered in the phase 3 cabozantinib trial . Fistulas in the gastrointestinal tract are rare (<1%), although extremely dangerous, adverse events associated with cabozantinib. Practical guidelines on the management of the more common adverse events have been developed (Rao & Cabanillas 2018, Cabanillas & Takahashi 2019. Patients need to be carefully assessed, focusing on their performance status along with the burden and location of metastases. After treatment commencement, regular review is required to allow for the early and aggressive management of potential adverse events. In cases of less serious adverse events (grade 1-2), dose modification, by lowering the administered dose, should be considered in parallel with symptomatic treatment; however, in cases of serious adverse events (grades 3-4), treatment should be interrupted or suspended.
In the case of progression on an initial TKI, treatment with another TKI, besides those currently approved, may be considered when accessible. Lenvatinib is a TKI inhibiting VEGFR-1, 2, and 3, fibroblast growth factor receptor (FGFR)-1-4, platelet-derived growth factor receptor (PDGFR), RET and KIT signalling pathways, that has been approved for the treatment of differentiated thyroid cancer, but only scanty data exist on its use in MTC. In a phase 2 study including 59 patients with unresectable progressive MTC, disease control was achieved in 80% (95% CI: 67-89%), 44% of whom had stable disease . Median PFS was 9.0 months (95% CI: 7.0 -not evaluable). Similarly, in a smaller but more recent phase 2 trial of 51 patients with all histological types of progressive thyroid cancer, including 9 patients with MTC, the median PFS for MTC patients was 9.2 months, with a 22% response rate . Lenvatinib can induce posterior reversible encephalopathy syndrome, which in combination with Takotsubo cardiomyopathy resulted in a fatal outcome in one case (Chae et al. 2018). Other rarer adverse events include thrombocytopaenia, fistula formation and heart failure (Cabanillas & Takahashi 2019).
Sorafenib is a dual inhibitor, targeting Raf pathways as well as tyrosine kinase pathways including VEGFR-1, VEGFR-2, VEGFR-3 and PDGFR-β. In two meta-analyses including 101 and 99 progressive MTC patients, respectively, from 8 studies using sorafenib as the single TKI treatment (previous treatments included surgery,
Therapy of metastatic medullary thyroid cancer
In the first meta-analysis, an overall partial response and stable disease were found in 21 and 58% of MTC patients, respectively (Vuong et al. 2019), while in the other, an objective response (including complete and partial responses) was found in 27.6% of patients with a PFS of 12.4 months (95% CI: 8.4-16). However, this was associated with a high discontinuation rate of up to 32.3% (95% CI: 24.3-40) of patients due to toxicity (Vuong et al. 2019, Efstathiadou et al. 2021. Responses to sorafenib were not durable, with resistance developing, usually within 1-2 years (Vuong et al. 2019).
Anlotinib is a TKI that inhibits both tumour angiogenesis and tumour cell proliferation by blocking VEGFR, FGFR, PDGFR and c-Kit simultaneously (Shen et al. 2018, Li et al. 2021. In a single-arm phase 2 study conducted in China, 58 patients with progressive MTC (unresectable, locally advanced or metastatic MTC) treated with anlotinib exhibited an objective response rate (ORR) of 56.9% (Sun et al. 2018), with a PFS rate at 48 weeks of 85.5%. The most common adverse events were palmarplantar erythrodysaesthesia, hypertriglyceridaemia, cholesterol elevation, fatigue and proteinuria.
In a recent meta-analysis including 33 studies (23 prospective, 9 retrospective, 1 observational) with 99 metastatic MTC patients and 16 cohorts showing progressive disease ab initio, overall stable disease was observed in 46.2% of TKI-treated patients and progressive disease in 22.9% (Efstathiadou et al. 2021). In particular, the use of TKIs conferred a PFS of 23.3 months (95% CI: 21.07-25.5). The meta-analysis included patients treated with various TKIs including axitinib, cabozantinib, dovinitib, imatinib, lenvatinib, motesanib, pazopanib, sorafenib, sunitinib, vandetanib and combinations of these treatments. In particular, progressive disease occurred in 23.7% of patients treated with vandetanib, in 22.6% treated with cabozantinib and in 19.3% treated with sorafenib. Grade 3 or greater adverse events occurred in 48.5% of TKI-treated patients and drug discontinuation was reported in 44.7%. Vandetanib induced an objective response in 33.8% and cabozantinib in 27.7% of MTC patients (Efstathiadou et al. 2021).
A limitation of TKI treatment is that the tumour cells eventually develop resistance, associated with tumour progression, independent of the type of TKI used or molecular tumour profile. Secondary mutations in the kinase domains that inhibit the binding of TKIs, usually downstream from the TKI target, resulting in a mechanism that bypasses the action of the drug (Viola et al. 2016). In MTC patients, acquired RET V804M 'gatekeeper' resistance mutations have been described in vandetanib-treated patients (Subbiah et al. 2018a). In such cases, a second TKI such as cabozantinib that acts via more than one molecular pathway might be considered in order to counteract such resistance (Priya et al. 2017). Of note, the discontinuation of TKI treatment could lead to a rapid increase in tumour growth and disease progression (Resteghini et al. 2017), presumably because the TKI blockade was being partially (but only partially) reversed by the activation of alternative signalling pathways. In such a situation, combination drug use may be considered, although this approach is investigational (see below).

Specific inhibitors of RET oncogene alterations
In 2020, two new RET-selective inhibitors were FDAapproved: selpercatinib (LOXO-292) and pralsetinib (BLU-667), with the former now also having been approved by the EMA.
Recently, selpercatinib (LOXO-292), which selectively targets oncogenic RET alterations, was approved by the FDA in the United States for the treatment of RET-mutant MTC and RET fusion thyroid cancers (Wirth et al. 2020); it has recently also been approved in the United Kingdom in patients resistant or intolerant of the multikinase inhibitors. In 55 RET-mutant MTC patients who had previously progressed through or not tolerated vandetanib and/or cabozantinib, the objective response rate was 69% (95% CI: 55-81), and the 1-year PFS was 82% (95% CI: 69-90). In 88 RET-mutant MTC patients who had not previously received vandetanib or cabozantinib, the response rate was 73% (95% CI: 62-82), and the 1-year PFS 92% (95% CI: 82-97). The most common adverse events of grade 3 or greater were hypertension in 21% of patients, increased alanine aminotransferase (ALT) and aspartate aminotransferase (AST) levels in 11 and 9% respectively, hyponatraemia in 8% and diarrhoea in 6%. Similar response rates were observed in 3 out of 11 patients (27%) who harboured the RETV804M mutation, which confers acquired resistance to vandetanib, showing clinical and radiological responses with a maximum change in tumour size ≥60% after receiving selpercatinib, although resistance eventually developed (Wirth et al. 2020).
In 2020, both the new generation TKIs, selpercatinib and pralsetinib, were granted FDA approval in patients with advanced RET-mutant MTC who require systemic therapy and indeed they are considered a highly potent treatment in patients with RET-driven MTCs (Subbiah et al. 2018a). However, there is concern that these highly selective inhibitors may not be as 'durable' as the multikinase inhibitors, with resistance being seen to occur more rapidly as follow-up studies progress , Lin & Gainor 2021.
The vast majority of MTC cases resistant to selpercatinib or pralsetinib are RET-mutant negative; further investigation of RET-independent resistance mechanisms has demonstrated that the resistance to selective RET inhibition may be driven by acquired MET or KRAS amplifications . In the cases of resistance in which the RET pathway is involved, RET solvent front (G810) mutations as well as hinge region (Y806) mutations confer selpercatinib resistance through steric hindrance with drug binding , Solomon et al. 2020, Subbiah et al. 2021. Moreover, using X-ray crystal structures of RET-selpercatinib and RET-pralsetinib complexes, these inhibitors are not susceptible to RET gatekeeper (V804) mutations because of their mode of binding, while remaining susceptible to nongatekeeper mutations (Subbiah et al. 2021). In addition, gatekeeper mutations have been more readily identified after progression on approved TKIs in EGFR-mutant or ALK fusion-positive non-small-cell lung cancer (NSCLC) as compared with RET fusion-positive NSCLC ). It will be necessary to develop RET inhibitors with maintained potency against RET resistance mutations, and it may be that combination strategies will be needed to effectively overcome resistance in these patients. As an example, TPX-0046 is a potent RET/SRC inhibitor which has demonstrated preclinical potency against RET G810 solvent front mutations and is currently in phase 1 testing in patients with advanced RET-altered solid tumours (NCT04161391).

When to initiate TKIs and what TKI to choose
The initiation of systemic treatment with TKIs is recommended in the setting of progressive or extensive distant metastases. The current recommendation (Wells et al. 2015, Haddad et al. 2018 is to consider treatment for tumours >1-2 cm in diameter growing >20% per year, or for symptomatic control: clearly depending on availability, one could use the selective RET-antagonists for germline or somatic RET-mutation-positive tumours and vandetaninib or cabozantinib when these are absent. However, in the United Kingdom, the more selective RET inhibitors are currently only funded when patients progress through the multikinase inhibitors or when these cannot be tolerated, even after dose reduction. It may be that the superior adverse-event profile of pralsetinib and selpercatinib will determine their first-line use, availability and expense allowing, but as yet there is a question mark over their response durability, and longer trial follow-up is awaited. It is also unclear as to whether the presence of a RET mutation, either germline or somatic, is mandatory before the use of either the multikinase or the RET selective inhibitors. TKIs in MTC patients with progressive disease are associated with a moderate therapeutic benefit, including disease stability or partial responses in up to 73% of cases, decreasing the 10-year disease-specific mortality rate from 38 to 13.5%, mainly seen in patients harbouring RET M918T mutations (Dvorakova et al. 2008, Taccaliti et al. 2011, Kuo et al. 2018. While the toxicity of these drugs is not negligible, it is usually manageable.

Combined targeted therapies
Certain small molecule inhibitors possess efficacy against MTC in vivo in combination with other drugs (Priya et al. 2017). Sunitinib and cisplatin cooperatively interfere with the autophagic lysosomal pathway, and this combination appears to be active in metastatic or progressive MTC (Lopergolo et al. 2014). Acquired resistance to targeted therapy could possibly be overcome by employing a combination of targets such as everolimus plus 5-aza-2′deoxycytidine (AZA), as this combination was found to have a synergistic effect even in everolimus-resistant cell lines (Vitale et al. 2017). In other neuroendocrine tumours, combination therapies may show synergistic effects, at least in vitro (Aristizabal Prada et al. 2018).
Therapy of metastatic medullary thyroid cancer

Monotherapy agents and epigenetic therapy
Nelfinavir (NFV), a heat shock protein (HSP)-90 chaperone, is required for the folding and stability of RET mutants (Alfano et al. 2010). HSP-90 overexpression is found in a significant proportion of MTCs and may correlate with metastatic potential and RET mutations. HSP90 is a molecular target for the HIV-protease inhibitor NFV. It has been shown to have a wide spectrum of activity in vivo against MTC cells and may emerge as an important therapeutic option (Valea & Georgescu 2016).

mTOR inhibitors
Given that RET and RAS activate the PI3K/AKT/mTOR pathway, the anti-proliferative activity of everolimus, an inhibitor of mechanistic target of rapamycin (mTOR) approved for the treatment of neuroendocrine tumours and renal cell carcinoma, was demonstrated in an MTC human cell line (Grozinsky-Glasberg et al. 2010) and some activity in a single patient with an MTC (Druce et al. 2012). In a small phase 2 trial, seven patients with progressive metastatic or inoperable MTC (Schneider et al. 2015) were treated with everolimus. The median PFS was 33 weeks, but no objective response was observed (Schneider et al. 2015). Similar findings were encountered in another phase 2 trial that included nine MTC patients treated with everolimus (Lim et al. 2013). The median PFS was not reached (after a median follow-up of 11 months), indicating that everolimus may have activity against MTC and that larger prospective studies are warranted. Biochemical response, defined as calcitonin and CEA levels ≥50% lower than baseline, was observed in three (30%) and four (44%) patients, respectively. The most common treatment-related AEs were mucositis, anorexia and AST/ALT elevation, and were in general mild to moderate (Lim et al. 2013). Nevertheless, promising experimental data have been reported in MTC-1.1 cells exposed to combination treatment including everolimus and sorafenib or sunitinib (Lin et al. 2012). Everolimus increased the antiproliferative effects of sunitinib and sorafenib by 24 and 27%, respectively, in MTC-1.1 cells and by 20 and 23%, respectively, in thyroid papillary (TT) cells (Lin et al. 2012).
In another report, everolimus was prescribed in addition to vandetanib in a patient who developed disease progression, resulting in a 25% reduction of tumour volume for 8 months (Heilmann et al. 2016). The combination of RET kinase inhibitors and mTOR inhibitors might be an interesting dual-targeting strategy, but further evaluation in larger clinical trials is mandatory for establishing the most appropriate therapeutic regimen. Most importantly, such combinations, while theoretically attractive and soundly-based, may lead to a compounding of adverse effects rendering them clinically impractical.

Immunotherapy
Over the past few years, immunotherapy has been established as an oncological treatment for several types of malignancies, and preclinical studies on MTC have revealed potential benefit (Naoum et al. 2018). Data on clinical studies in MTC are scarce, but in a single phase 1 study including a patient with MTC, blocking the PD-1/PD-L1 interaction using nivolumab, an anti-PD-1 agent, resulted in a partial response (Yamamoto et al. 2017). Another case report also described a patient with MTC already treated with sunitinib and, having undergone a 3-month trial with the GI-6207 cancer vaccine, was enrolled in a phase 1 trial with the PD-L1 inhibitor avelumab (NCT01772004). Following treatment, he developed stable disease along with a >40% decrease in his calcitonin level (Del Rivero et al. 2020). Very preliminary results of a phase 2 trial (NCT03246958) evaluating nivolumab plus ipilimumab in seven patients with progressive MTC and prior TKI failure showed a lack of objective response in all seven patients, although without providing any further specific detail .
It is well known that high tissue tumour mutational burden (TMB) is a useful marker to recognise patients with recurrent or metastatic advanced solid tumours who may show a better therapeutic benefit from immune checkpoint inhibitors. Therefore, the NCCN guidelines for MTC management suggest that pembrolizumab should be considered in patients with symptomatic, progressive MTC with high TMB ≥ 10 mutations/megabase (NCCN 2021). MTC is reported to show low PD-L1 expression in both tumour cells and tumour-infiltrating immune cells and no microsatellite instability (Bongiovanni et al. 2017, Bi et al. 2019, irrespective of the presence or absence of either desmoplasia, lymph node metastases and/or RET mutation , Shi et al. 2019, Ingenwerth et al. 2020). Nevertheless, a cohort of 201 consecutive Chinese patients with MTC (Shi et al. 2019) demonstrated that PD-L1 positivity was associated with more aggressive clinicopathologic features, such as larger tumour size, number of Endocrine-Related Cancer lymph node metastases and advanced TNM staging and was independently predictive of morphological and/or biochemical recurrence or persistent disease (Sørlie et al. 2003). Similarly, PD-L1 positivity found in approximately 20% of surgical specimens (although very faint or moderate staining) significantly correlated with distant metastases at surgery (Bi et al. 2019). The correlation between PD-1/ PD-L1 status and prognosis suggests that immunotherapy targeting PD-L1/PD-1 might indeed be effective in treating advanced MTC under some circumstances. Five registered clinical trials are currently active studying immune-checkpoint inhibitors in progressive MTC patients (Table 2). Pembrolizumab (NCT03072160), nivolumab and ipilimumab (NCT03246958) are currently being evaluated in phase 2 trials for the treatment of recurrent or metastatic MTC. Interestingly, among the registered studies, there are also trials evaluating the combination of TKI agents together with immunotherapy or chemotherapy as a novel therapeutic option in the treatment of advanced MTC, although the combination of immunotherapy and chemotherapy is well established in other areas of oncology (Table 2).

Peptide receptor radionuclide therapy (PRRT)
Several studies targeting somatostatin receptors (SSTRs) with radionuclides have been reported in patients with progressive, inoperable or advanced metastatic MTC (Grossrubatscher et al. 2020), using radiopharmaceuticals such as 90 Yttrium ( 90 Y), 177 Lutetium ( 177 Lu) or 111 Indium ( 111 In) (Salavati et al. 2016, Parghane et al. 2020, Satapathy et al. 2020. In a recent review including 11 studies and 117 MTC patients treated with 90 Y, 177 Lu or 111 In, radiological response data showed that 37.6% of patients developed progressive disease, 54.7% stable disease, 5.1% partial response and 2.6% complete response; however, response criteria across the studies were heterogeneous (Grossrubatscher et al. 2020). Besides radiological evaluation, biochemical responses based on calcitonin plasma concentrations showed that 37% of the patients had a more than 50% calcitonin decrease (Grossrubatscher et al. 2020). In a prospective phase II study (n = 31 patients ), a prolongation of calcitonin doubling-time levels in 58% of patients post-90 Y-DOTATOC was reported (Iten et al. 2007). In general, patients with partial responses, disease stability and a biochemical response showed prolonged survival (Iten et al. 2007, Salavati et al. 2016, Parghane et al. 2020. After 17 years of experience following treatment with 177 Lu-DOTATATE (Beukhof et al. 2019), albeit in a very small group of patients, it was concluded that this treatment could be considered in patients with high uptake on an 111 In-DTPA-octreotide scan and positive somatostatin receptor-2a (SSTR2a) expression in tumour samples (Beukhof et al. 2019). Twenty-one patients with advanced MTC treated with PRRT ( 90 Y-DOTATATE, 90 Y-DOTATOC, 177 Lu-DOTATATE or 177 Lu-DOTATOC) achieved a median time-to-treatment-failure interval of 14 months (95% CI: 8-25) and a median OS of 63 months (95% CI: 21, not reached). Predictors of poorer OS included a short calcitonin doubling-time (≤24 months), strong 18 F-FDG avidity and age ≥60 years (Hayes et al. 2021). A larger retrospective series of 43 patients all treated with 177 Lu-DOTATATE showed a RECIST partial response in 4%, stable disease in 58% and a median PFS of 24 months (Parghane et al. 2020).
PRRT is often well tolerated but is not free of adverse effects such as nausea, asthenia and elevation of liver enzyme levels, although they are usually mild. Serious complications, such as radiation nephrotoxicity and treatment-related myeloid neoplasms, are relatively rare especially with the more frequently used 177 Lu-labelled radiopeptides (Brabander et al. 2017, Parghane et al. 2020, Sonbol et al. 2020. Adverse events of all grades of nephrotoxicity were reported in 0-1.3% of patients, with grade 3-4 in 0.8%, and grade 3-4 haematological toxicity in up to 6.5% of cases (Grossrubatscher et al. 2020, Parghane et al. 2020. Proximal tubular reabsorption of the radiopeptide leads to glomerular fibrosis, which is particularly observed after treatment with 90 Y-DOTATOC (Parghane et al. 2020), and thus the 177 Lu isotope is usually preferred over 90 Y. Cumulative and per-cycle renal uptake dose, age, presence of hypertension, diabetes mellitus and previous chemotherapy with nephrotoxic agents could potentially accelerate the decrease in renal function after PRRT. Considering these risk factors, one can modify the treatment plan or change the choice of radiopeptide based on the overall tumour burden (Rottenburger et al. 2020).
The use of a concomitant radiosensitising agent, such as low-dose oral capecitabine, with 177 Lu-DOTATATE, might be also beneficial, as suggested by recent studies reporting up to an 86% disease control rate (Grossrubatscher et al. 2020, Satapathy et al. 2020. However, the only prospective, randomised study (CONTROL NETs), though noncomparative, of PRRT with radiosensitising chemotherapy (capecitabine/temozolomide) in metastatic midgut NETs showed a similar PFS rate at 15 months compared to PRRT alone (Pavlakis et al. 2020). The objective response rate was numerically higher in the combination arm but at the expense of greater G3/4 toxicity.
There is recently in vitro and in vivo evidence that somatostatin receptor subtype-2 (SSTR2) antagonists Endocrine-Related Cancer Endocrine-Related Cancer are better tools to target neuroendocrine tumours than SSTR agonists (Reubi et al. 2017 Endocrine-Related Cancer approximately 3.5 times higher than those obtained with 177 Lu-DOTATATE (Wild et al. 2014). While meta-iodobenzylguanidine (MIBG) labelled with 123 Iodine is most effectively used in diagnostic scintigraphy, radio-labelled 131 I attached to MIBG is mainly used as a therapeutic, with uptake found in up to 35% MTC (Pasieka et al. 2004). Concerning the efficacy of 131 I-MIBG treatment in MTC, data are limited to case reports and evidence from a prospective randomised trial is lacking (Mukherjee et al. 2001, Sze et al. 2013). This radionuclide may in general be more myelotoxic (Sze et al. 2013). CEA targeting using murine or chimeric anti-CEA bispecific antibodies and pre-targeted hapten-peptides labelled with 111 In or 131 I has been also developed for radioimmunotherapy in relapsed MTC (Bodet-Milin et al. 2016) ( Table 2).
The cholecystokinin-2 receptor (CCK2R or gastrin receptor) is over-expressed in several tumour types, including MTC. CCK2R targeting shows promise for MTC imaging and represents, together with SSTR ligands, potential theranostic approaches in selected cases (Refardt et al. 2021). Many peptide-based radiopharmaceuticals for targeting CCK2R have been developed (Behr & Béhé 2002). In a recent study, the 177 Lu-labelled minigastrin analogue 177 Lu-PP-F11N showed promising results in six patients with advanced MTC (Rottenburger et al. 2020). Recently, a new highly metabolically stable minigastrin analogue (DOTA-MGS5) has shown excellent targeting properties in preclinical studies, supporting its possible translation into clinical trials (Klingler et al. 2019).

Other systemic or local therapies Chemotherapy
The objective response rate of MTC patients to classical cytotoxic chemotherapy such as dacarbazine and 5-fluorouracil is 10-20%, with most studies performed on limited numbers of patients and without robust evaluation criteria such as RECIST (Hadoux & Schlumberger 2017). Long-term responses are uncommon.

External beam radiotherapy
External beam radiotherapy (EBRT) may be recommended to improve loco-regional control in patients at high risk of relapse in the neck. The usual indication for EBRT is macroscopically unresectable tumour in a patient older than 45 years of age. EBRT is indicated to improve the local control of disease in cases of local recurrence or loco-regional lymph node metastases, or as palliative therapy to reduce pain from bone metastases or to treat brain metastases (Wells et al. 2015). A few studies have shown an improvement in locoregional control but no survival benefit, confirming that although neck disease can be controlled in high-risk patients, the progression of distant disease is still a significant problem (Schwartz et al. 2008, Martinez et al. 2010, Brierley & Sherman 2012, Call et al. 2013, Tsoukalas et al. 2017. Localised EBRT and/or bisphosphonate administration should be considered for painful bone metastases or for the prevention of skeletalrelated events (Farooki et al. 2012). Recommendations from a group with extensive clinical experience suggest quarterly zoledronic acid (Haugen et al. 2016). According to ATA guidelines for MTC, adjuvant EBRT can be considered in patients with a high risk of recurrence or with incompletely resected MTC (Haugen et al. 2016). However, EBRT was not associated with improved overall or disease-specific survival in a recent study of patients with MTC (Jin et al. 2021).

Embolisation or cryoablation
Embolisation or cryoablation as a palliative therapy may be beneficial in select cases to decrease tumour burden, pain or refractory diarrhoea associated with liver metastases (Fromigué et al. 2006). Radiofrequency thermo-ablation is frequently applicable to bone, liver and lung to treat single metastases or a single progressive and symptomatic lesion in the context of stable disease (Eisele 2016). Another local treatment is the conventional transarterial chemoembolisation (TACE) or radioembolisation (TARE), sometimes used in advanced cases of liver metastatic disease, especially when liver metastases are smaller than 3 cm and liver involvement is less than 30% (Roche et al. 2003).

Hormonal syndromes secondary to MTC
MTC has also been associated with the ectopic production of bioactive compounds, peptides or amines, leading to distinct clinical syndromes (Kaltsas et al. 2010). The most commonly described ectopic syndrome is ACTH leading to Cushing's syndrome that in some cases can be severe. Other compounds include serotonin, somatostatin, pro-opiomelanocortin, vasoactive intestinal peptide, chromogranin A, amyloid, gastrin-releasing peptide and histamine (Wells et al. 2015). Patients with advanced MTC may experience debilitating diarrhoea and flushing due to hypersecretion of calcitonin, VIP, or increases in intestinal Endocrine-Related Cancer motility (Cox et al. 1979, Rambaud et al. 1988. Antimotility agents (loperamide or codeine) may be initially used for symptom relief. Further options for refractory cases include somatostatin analogues, although tachyphylaxis typically occurs over time (Zatelli et al. 2006). For patients with extensive liver metastases, surgical resection, percutaneous radiofrequency ablation or arterial (chemo)embolisation may be considered in an attempt to reduce the calcitonin levels (Fromigué et al. 2006). Cushing's syndrome due to ectopic ACTH secretion is found in 0.7% of MTC patients, accounting for 2.2-7.5% of all ectopic ACTH-Cushing's syndrome cases (Barbosa et al. 2005). Fifty cases of MTC-related Cushing's syndrome have been reported so far. In those cases, the tumour cells are positive for calcitonin but mostly display negative immunostaining for ACTH, presumably due to the high secretion rate (Shahani et al. 2010). Ectopic Cushing's syndrome from MTC is associated with significant morbidity and mortality, as secondary complications of hypercortisolism account for 50% of the mortality in these patients (Hayes & Grossman 2018). In the past, management of hypercortisolism was limited to decreasing metastatic tumour burden and using anti-adrenal therapies such as ketoconazole, mitotane and/or metyrapone. Surgical intervention with bilateral adrenalectomy is another option available to some patients (Hayes & Grossman 2018). However, systemic therapy with tyrosine kinase inhibitors such as vandetanib offers a further management strategy for disease control of metastatic disease and might now be considered a first-line therapy for ectopic Cushing's syndrome, following control of the excess cortisol, in the setting of unresectable disease or progressive, metastatic disease, as well as for controlling secretory Cushing's syndrome (Wells et al. 2012, Paepegaey et al. 2017.

Future perspectives
In a subset of MTC, octreotide and pasireotide inhibit cell proliferation and invasiveness, especially when the expression of SSRT2 and SSRT5 is high, supporting their potential use in the control of tumour growth. On the contrary, MTC with high SSRT3 expression was unresponsive to both octreotide and pasireotide (Zatelli et al. 2006). Calcitonin secretion was not affected by octreotide or by pasireotide in primary cultured MTC cells, regardless of their responsiveness to the SSA's antiproliferative effects. These data confirm a dissociation between the anti-mitotic and anti-secretory effects of SSAs in primary cultured MTC cells (Zatelli et al. 2006).
Activating transcription factor 4 (ATF4) is a negative regulator of the RET tyrosine kinase receptor in MTC. Low ATF4 protein or ATF4 loss alone had a significant negative impact on median survival compared to high protein expression or diploid ATF4. The combination of somatic RET expression and low ATF4 protein levels further decreased overall survival. These data suggest that ATF4 may predict response to tyrosine kinase inhibitors, serve as a prognostic marker for personalised care and represent a therapeutic target in MTC (Williams et al. 2021). The small-molecule ONC201 caused apoptotic MTC cell death, decreased transcription of RET, VEGFR2, IGFBP2 and increased mRNA levels of ATF4 representing a promising strategy for the treatment of all patients with MTC, including those with TKI-refractory disease and other cancers with RET abnormalities (Bagheri-Yarmand et al. 2021).
Many novel therapeutic strategies are being pursued to develop personalised treatment approaches in MTC. Inhibitors of alternative targets and pathways, such as aurora kinase inhibitors, mTOR inhibitors, farnesyltransferase inhibitors and exploiting microRNA levels, seem promising (Gild et al. 2013, Joo et al. 2019). The potential for administering these drugs in a neoadjuvant setting to facilitate surgery and aiming for surgical cure are also on the horizon . Recently, new targets have been highlighted, such as the glucose-dependent insulinotropic polypeptide receptor (GIPR) (Gourni et al. 2014) and minigastrin, a ligand targeting cholecystokinin-2 (CCK2) receptors, as noted previously. Ongoing studies (under recruitment or active) are presented in Table 2.

Conclusions
The introduction of the multikinase inhibitors, together with the recent more selective RET inhibitors, has opened a new phase in our treatment of these rare but fascinating tumours. While current response durations are relatively limited, they do suggest the prospect of a gradual improvement in therapies over time, with increasing durability as our understanding of the molecular biology develops. Nevertheless, their efficacy should be further investigated by larger trials with robust methods. In time, we would anticipate a more personalised therapeutic stratagem based on individualised patient tumour profiling and selective signalling pathway targeting. Treatments reducing tumour bulk, such as additional surgery and (chemo)embolisation, as well as external beam radiotherapy, are also utilised, but, generally, the results
Therapy of metastatic medullary thyroid cancer